[PLUG] experiments - OT: FiOS auxiliary power for BBU

[wes]

On Fri, Feb 19, 2010 at 10:42 AM, Tim Wescott <tim at wescottdesign.com> wrote:

> Keith Lofstrom wrote:
> > On Fri, Feb 19, 2010 at 10:01:25AM -0600, Fred James wrote:
> >
> >> "but not with a 0V battery" - I don't know - from the documentation one
> >> might expect the BBU to shutdown leaving some "reserve" power in the
> battery
> >> Now, Verizon "tech" did say the aux port wasn't hooked up, but APC
> >> couldn't understand that statement - i.e.: APC's documents contradict
> >> Verizon's statement - but I don't know how to verify that either way?
> >>
> >
> > Verizon is not in the business of telling the truth.  They are in the
> > business of keeping their equipment safe from hardware hackers like us.
> > Obviously, there are risks, starting with some clown hooking a car
> > battery backwards into that port.  If I was programming an Indian call
> > center, I would have them say whatever was necessary to keep people like
> > us from fooling with the customer equipment.  I'm grateful that they
> > were willing to help me plug into the ethernet port on the ONT, rather
> > than route through the Actiontec cable modem as originally installed
> > (another story).  They could have told me to jump in a lake.
> >
> > The aux port does hook up to the aux indicator light, and there is a
> > 12.5V threshold, so it is not just a matter of an otherwise-unconnected
> > LED and resistor.  I imagine that under the right circumstances, the
> > port will provide power though the BBU.
> >
> > I'm guessing that replacing the good battery with an almost-but-not-quite
> > discharged one will cause the main BBU to draw power from the aux port.
> > The BBU control circuitry that does the switching may be powered off the
> > main internal battery, drawing a trickle of current.
> >
> > If the external APC unit has a charger, then that is good news, because
> > it means the main APC unit is not trying to charge it, and probably
> > does not care how big the external battery is.  Car batteries are much
> > cheaper per watt-hour than the small gel cells, and you can keep them
> > on a trickle charger.  In extremis, you could even power the unit from
> > the battery in a car, BUT DO NOT CONNECT IT WHEN THE CAR IS RUNNING,
> > the ignition spikes would likely kill the APC BBU.
> >
> >
> If you're looking for inexpensive lead-acid batteries for this service,
> look to deep discharge marine, RV or golf cart batteries.  Car batteries
> are optimized to generate LOTS of current for a short time, and are
> presumed to rarely get deeply discharged.  This is because that current
> is needed, and they generally still have considerable remaining capacity
> when they're too 'dead' to turn an engine.  As a consequence of being
> optimized for low price and high current output they don't do a good job
> controlling where the material plates on during charge after a deep
> discharge, so they tend to suffer an early death if they go through too
> many deep discharge/charge cycles.
>
> There is considerable collected wisdom on this subject in the amateur
> radio literature, from folks building emergency-power stations suitable
> for the aftermath of fire, flood, and earthquake.  They all say "don't
> use car batteries".
>
> Anything called "deep discharge" will (well, should) be designed to be
> deeply discharged -- and recharged -- on a regular basis, and will
> probably severely compromise cold cranking amps while doing so.
>
> --
> Tim Wescott
>

The more common term that I've seen used is "deep cycle."

-wes